**1. Introduction**

## **1.1 Cell membranes**

The cell membrane is the structure that gives cells their individuality by separating them from the extracellular medium and from other cells. It regulates the transport of ions, molecules and signals towards the interior and exterior of the cell. Membranes

confer the selective permeability that maintains the differences in composition between the cytoplasm and the extracellular medium, which in turn regulates the cellular volume and the cellular response to the different signals it receives or generates. In eukaryotes, membranes divide the cell interior into compartments, or organelles [1]. Membranes are made of proteins, lipids, and carbohydrates. Proteins and lipids represent almost its entire composition, while carbohydrates represent less than 10% [2]. In particular, lipids represent from 25 to 80% of the membrane weight, and they belong to three chemical groups: phospholipids, glycolipids and sterols. All membrane lipids are amphipathic molecules that have an hydrophilic region and an hydrophobic region. Phospholipids and glycolipids spontaneously assemble into closed bilayers in aqueous mediums and constitute the membrane matrix [1, 3].

#### **1.2 Molecular shape of membrane lipids and their molecular associations**

X-ray diffraction studies of membrane lipids have shown that the area of their polar regions (AO) significantly varies in comparison with the cross-sectional area of their hydrocarbon chains (AH). These studies have revealed three molecular shapes: cylindrical (**Figure 1A**), conical (**Figure 1B**) and inverted conical (**Figure 1C**) [4]. In cylindrical lipids, the area of the polar region is almost equal to the cross-sectional area of the non-polar region, so the AH/AO ratio is close to 1. In conical lipids, the area of the polar region is less than the cross-sectional area of the non-polar and AH/AO is >1; while in the inverted conical lipids this relationship is inverse and AH/AO is <1 [5]. Cylindrical lipids, such as phosphatidylcholine (**Figure 1D**), phosphatidylglycerol, phosphatidylinositol, phosphatidylserine, and sphingomyelin, associated in aqueous medium to form closed bilayers or liposomes (**Figure 1E**), and represent 60 to 70% of membrane lipids. Conical lipids, such as phosphatidate (**Figure 1F**), phosphatidylethanolamine, diacylglycerol, and cardiolipin, assemble into an hexagonal phase II (**Figure 1G**), which consists of hexagonally packed cylinders, with the polar regions directed towards the interior of the cylinder where they form an aqueous pore of around 50 Å in diameter [5]. Inverted conical lipids, such as lysophospholipids and gangliosides (**Figure 1H**), assemble into a micellar phase (**Figure 1I**), with the polar regions towards the outside and the non-polar regions towards the inside. The membrane has lipids with the three molecular forms, and the higher proportion of cylindrical lipids compared to conical and inverted conical lipids allows the association of these three kinds of lipids into bilayers. However, membrane lipids can also have different molecular associations. 31P NMR studies have indicated the presence of lipids associated in non-bilayer phospholipid arrangements (NPA) (**Figure 1J**) in membranes with high metabolic activity, such as in cancer cells and in rat liver microsomes [6].

#### **1.3 Supramolecular organization of cell membranes**

Singer and Nicolson proposed the fluid mosaic model for cell membranes in 1972. In this model, the integral proteins are inserted in the lipid bilayer, which constitutes the mosaic, and it is fluid because the interactions between lipids and between lipids and proteins are non-covalent, which allows these molecules to move laterally across the membrane.

In cell membranes, the role of lipids is mainly structural, and the type of fatty acids they contain determines the fluidity of the membrane. The hydrophobic effect *Anti-Non-Bilayer Phospholipid Arrangement Antibodies Trigger an Autoimmune Disease… DOI: http://dx.doi.org/10.5772/intechopen.106373*

#### **Figure 1.**

*Classification of membrane lipids based on their molecular shape. The lipid molecular shape (A–C) depends on its chemical structure (D–F). In addition, the three-dimensional association of lipids in an aqueous medium (G–I) depends on the lipid molecular shape. Diagram of lipids associated in a non-bilayer phospholipid arrangement (J).*

is the main force that maintains the organization of proteins and lipids in this model. The thickness of cell membranes, with their peripheral and integral proteins, is around 100 Å [7, 8].

Although the fluid mosaic model explains multiple properties of cell membranes, it does not consider the cylindrical, conical and inverted conical molecular shapes of lipids, nor their functional role. Cullis and colleagues proposed the metamorphic mosaic model, where the bilayer is formed by lipids of the three molecular forms, and may temporarily have lipid associations different from the bilayer, such as NPA (**Figure 1J**). These structures, which form a microdomain, may participate in many cellular functions, including phagocytosis, membrane junctions, transport of ions and polar molecules, membrane fusion in exocytosis and endocytosis, protein insertion, and formation of polar pores and compartments [4]. The importance of this model lies in proposing lipid associations different than the lipid bilayer, like NPA, which participate dynamically in membrane functions. Therefore, they attribute a functional role to lipids, in addition to the structural role of Singer and Nicolson.

Lipids are generally poorly immunogenic molecules [9]. From the two molecular associations that can occur in cell membranes, the lipid bilayer is considered to be the least immunogenic, because it constitutes the lipid matrix of all cell membranes. NPA are also poorly immunogenic, since they are transitory and therefore are not detected by the immune system; however, if they are stabilized by amphipathic molecules, an immune response is induced with the production of antibodies against these lipid structures [10]. Among the molecules that have been found to stabilize NPA are drugs that, as a side effect, induce a disease similar to Systemic Lupus Erythematosus (SLE) in humans. SLE is a chronic, multifactorial autoimmune disease with an unknown etiology. SLE patients present anti-nuclear, anti-histone, anti-cardiolipin and anti-DNA auto-antibodies that can form antigen-antibody complexes that damage multiple organs. This disease can affect the skin, joints, blood cells, kidneys and the nervous system [11]. The most common clinical manifestations are extreme tiredness, unexplained fever, skin rash, malar rash, and arthritis. Serious complications can also occur, such as lupus nephritis and autoimmune cytopenias [12]. SLE mainly affects women in a female to male ratio of 9:1 [13]. This disease can develop at any age; however, in most cases it occurs between the ages of 24 and 32 during the fertility peak, so female sex hormones are considered a key factor in the development of this disease [14]. According to the Lupus Foundation of America, at least 5 million people worldwide have lupus. There is a higher prevalence and incidence among the Hispanic, Asian, and African-American populations [15].

#### **1.4 Drug-induced lupus in humans**

Drug-induced lupus its generated by the chronic intake of certain drugs, which induce an immune response that triggers a disease that is very similar to, but less severe than SLE. There are about 38 drugs that cause drug-induced lupus, including hydralazine, procainamide and isoniazid, which are responsible for most of the cases [16]. The exact mechanism that leads to drug-induced lupus is not well understood; however, one factor that predisposes to its development is the rate at which drugs are metabolized, which is markedly decreased in patients with a genetic deficiency of N-acetyl transferase. These patients have a higher incidence of drug-induced lupus [17].

In addition, it has been reported that these lupus-inducing drugs can suppress central and peripheral tolerance, alter gene transcription in T and B cells, alter the balance and function of cytokines or their receptors, and modify the structure of

chromatin and self-antigens [17–19]. Another possible mechanism that would explain the involvement of these drugs in the development of lupus is the stabilization of NPA on cell membranes, which then become immunogenic. We have explored this mechanism in a mouse model of lupus induced by drug-stabilized NPA on liposomes [20].

#### **1.5 Mouse models of lupus**

Mouse models of lupus have been very important to understand the genetic, cellular and molecular mechanisms of this autoimmune disease [21]. B/W mice, MRL/*lpr* mice and BXSB mice have been the most frequently used mouse models of lupus. New Zealand Black (NZB) mice develop autoimmune hemolytic anemia in the early stages of their lives, with reticulocytosis, jaundice and splenomegaly. Anti-nuclear antibodies are found in these mice, although generally in low titers [21]. The cross between NZB mice and New Zealand White (NZW) mice produces B/W mice, which develop a more aggressive autoimmune disease than that of NZB mice; this disease has similar characteristics to human lupus. B/W mice have mutations in the major histocompatibility complex (MHC) genes, and present high titers of anti-nuclear and anti-DNA antibodies, and glomerulonephritis caused by immune complexes [22]. Females B/W mice are more severely affected than males [23].

Murphy-Roths large (MRL)/*lpr* mice have a mutation on the Fas gene, which leads to deficient B and T cell apoptosis. These mice present non-malignant lymphoid proliferation and manifestations of autoimmunity, including the production of anti-DNA and anti-ribonucleoprotein antibodies, glomerulonephritis, vasculitis, and arthritis [22]. BXSB/Yaa mice have a duplicated genome section that includes the toll-like receptor 7 (*Tlr7*) and the phosphoribosyl pyrophosphate synthetase 2 (*Prps2*) genes. These mice spontaneously produce anti-DNA antibodies and develop an immune complex-mediated glomerulonephritis that resembles the glomerulonephritis of SLE patients. In contrast with B/W mice and MRL/*lpr* mice, male BXSB/Yaa mice are more severely affected by the disease than females [24].

In B/W mice, MRL/*lpr* mice and BXSB mice, a genetic abnormality alters the regulatory mechanisms of the immune system and promotes the development of autoimmunity. In other mouse models, a lupus-like disease can be induced in mice that are not genetically susceptible to autoimmune disease. This category includes mice that develop lupus after receiving DNA/protein or RNA/protein complexes [25, 26], or after receiving pristane (2,6,10,14-tetramethylpentadecane) [27, 28], and it also includes the mouse model of lupus induced by drug-stabilized NPA on liposomes.

## *1.5.1 Mouse model of lupus induced by lipids associated in NPA*

This mouse model of lupus can be developed by the administration of liposomes bearing drug-stabilized NPA, or by the administration of the NPA-stabilizing drugs alone; these drugs include chlorpromazine (anti-psychotic), hydralazine (antihypertensive - diuretic) and procainamide (anti-arrhythmic) [20]. Mice develop IgG anti-NPA antibodies, followed by anti-cardiolipin, anti-histone, anti-nuclear and anti-coagulant antibodies. They present moderate alopecia, symmetrical facial lesions similar to those observed in SLE patients, and immune complex deposits between the dermis and the epidermis, in the walls of the glomerular capillaries and in the glomerular mesangium, as occurs in SLE. The symmetrical facial lesions and the deposition of immune complexes between the dermis and the epidermis are unique features of this mouse model that have not been described in other mouse models of lupus.

In this mouse model, drug-stabilized NPA become immunogenic and induce the production of anti-NPA antibodies, which are of the IgG class. Anti-lipid IgG antibodies have been detected in infectious diseases caused by mycobacteria [29] and in autoimmune diseases such as SLE [30]. For protein antigens, the development of IgG antibodies implies an isotype switch in germinal center reactions [31], but these events have practically not been studied *in vivo* for lipid antigens. Thus, this mouse model of lupus offers a unique opportunity to analyze the role of germinal centers, the extrafollicular reaction, and plasma cell development in response to lipid antigens.
